Trajectory Pattern Identification for Arrivals in Vectored Airspace

Grouping similar flight trajectories into a cluster, or a pattern, is an important data preprocessing step for data-driven methods in air traffic control as it affects the performance of applications such as separation assurance, anomaly detection and trajectory prediction. Especially in vectored airspace, it is challenging to obtain trajectory patterns as they are deeply embedded in the air traffic data. In this paper, we propose a clustering framework that can identify deeply embedded trajectory patterns in vectored airspace from the historical surveillance data using agglomerative hierarchical clustering and dynamic time warping. The proposed framework is illustrated with the Automatic Dependent Surveillance-Broadcast (ADS-B) data collected in the Incheon International Airport (ICN), South Korea.